Understanding complex crowd dynamics with generative neural simulators

作     者：Minartz, Koen Hendriks, Fleur Koop, Simon Martinus Corbetta, Alessandro Menkovski, Vlado 

作者机构：Department of Mathematics and Computer Science Eindhoven University of Technology Netherlands Department of Mechanical Engineering Eindhoven University of Technology Netherlands Department of Applied Physics and Science Education Eindhoven University of Technology Netherlands Eindhoven Artificial Intelligence Systems Institute Netherlands 

出 版 物：《arXiv》 (arXiv)

年 卷 期：2024年

核心收录：

主　　题：Stochastic systems 

摘      要：Understanding the dynamics of pedestrian crowds is an outstanding challenge crucial for designing efficient urban infrastructure and ensuring safe crowd management. To this end, both small-scale laboratory and large-scale real-world measurements have been used. However, these approaches respectively lack statistical resolution and parametric controllability, both essential to discovering physical relationships underlying the complex stochastic dynamics of crowds. Here, we establish an investigation paradigm that offers laboratory-like controllability, while ensuring the statistical resolution of large-scale real-world datasets. Using our data-driven Neural Crowd Simulator (NeCS), which we train on largescale data and validate against key statistical features of crowd dynamics, we show that we can perform effective surrogate crowd dynamics experiments without training on specific scenarios. We not only reproduce known experimental results on pairwise avoidance, but also uncover the vision-guided and topological nature of N-body interactions. These findings show how virtual experiments based on neural simulation enable data-driven scientific discovery. Copyright © 2024, The Authors. All rights reserved.